Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mrpulor-gh/nuphus --skill agent-orchestrationgit clone --depth 1 https://github.com/mrpulor-gh/nuphusWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mrpulor-gh/nuphus/agent-orchestration)<a href="https://agentmods.dev/skills/mrpulor-gh/nuphus/agent-orchestration"><img src="https://agentmods.dev/badge/skills/mrpulor-gh/nuphus/agent-orchestration/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mrpulor-gh/nuphus/agent-orchestration"><img src="https://agentmods.dev/badge/skills/mrpulor-gh/nuphus/agent-orchestration.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.07007 |
| Opus 5 | $0.00000 | $0.03504 |
| Sonnet 5 | $0.00000 | $0.01401 |
| Haiku 4.5 | $0.00000 | $0.00701 |
Grade A, and why
agent-orchestration scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- 终端类:curl 直调;**PowerShell 必须发 UTF-8 字节体**(`-Body ([Text.Encoding]::UTF8.GetBytes($json))`),字符串 body 会把中文压成 `?`(实测);纯 ASCII body 才可用单引号直传 How it starts
The opening of the file, as written. The whole thing — 350 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent 平台编排
外部 Agent(Claude Code / OpenCode / Hermes 等独立平台)协作指南。系统提示词已覆盖的(工具描述/annotation/Constitution)不重复,本技能只写外部 Agent 特有流程。
1. 登记(team.toml)
渐进登记:每用一个外部 Agent,在 plugin/team.toml 追加一段。只记稳定事实(mode/launch/process/window_hint/dispatch_steps/note),禁止记 PID/窗口句柄/坐标(坐标走 ui-maps;PID/hwnd 每次启动必变,hwnd 编号还会被 OS 复用给无关窗口——任何把易变事实固化进配置或缓存当派发依据的做法都是错的,进程管理职责归 Leader 当次实况)。
[opencode]
mode = "embedded" # background | embedded | standalone | web —— 决定交互协议(§2)
launch = "powershell -NoExit -Command opencode" # Leader 手动启动命令(见 §2 启动 SOP)
window_hint = "OpenCode" # 窗口标题特征(windows_list 匹配用)
process = "opencode.exe" # 进程名特征(process_list 识别依赖此字段)
双登记路径(实测均有效):
- 手改 team.toml:直接编辑保存即可被运行中系统即时读取(无需重启),适合快速实验与本轮临时接入;
- 配置中心录入:桌面 设置 → 外部 Agent → 新增(走
upsert_external_agent),适合正式归档——会联动生成 handoff 工作目录(read.md/memory.md/status.json),手改路径不生成这些文件(须后续补agent_init或由首次 dispatch 补建空骨架)。
调用决策:Read team.toml → process_list 看谁活着(活着优先)→ 按平台能力匹配。不维护历史评分。
跑通即归档(强制,当轮完成)
新外部 Agent 首次跑通验证后,必须当轮登记归档,禁止「以后再用再配」——不归档则下次会话无从知晓其存在与用法,跑通经验直接丢失。
- 配置中心录入:桌面 设置 → 外部 Agent → 新增(走
upsert_external_agent,字段填全):key:唯一 id(字母数字-_;保存后不可改)mode:background / embedded / standalone / web ——窗口分类依据(§2 交互方式由此决定)display_name/icon:状态栏与人读标识launch/args:启动命令与参数(Leader 手动启动时使用,§2 启动 SOP)window_hint:窗口标题特征(windows_list 匹配;终端类运行时标题常被覆写,hint 应选稳定前缀)process:进程名特征(process_list 识别依赖此字段)cooldown_secs/await_timeout_secs/timeout_action/confirm_keywords:启动冷却/超时动作/确认词表(缺省有默认值)description:职责一句话(路由提示;新 agent 自动同步为其.nuphus/handoff/{key}/read.md的职责段)note:Leader 专属实测备忘(如某热键不生效、某交互必须换路径等一手观察),随配置读取并在派发结果中回显;UI 禁止编辑
- 落盘核对:
plugin/team.toml出现该段且原有段未被破坏(写回是段级增量);新 key 联动生成 handoff 工作目录。 - 完整段示例(实测可用样本):
[opencode]
mode = "embedded" # background | embedded | standalone | web
display_name = "OpenCode"
icon = "terminal"
launch = "powershell -NoExit -Command opencode"
window_hint = "OpenCode"
cooldown_secs = 20
await_timeout_secs = 90
timeout_action = "screenshot_alive"
confirm_keywords = ["allow", "confirm", "proceed", "yes/no", "approve"]
# note = "实测备忘:…(Leader 专属,UI 不可编辑)
[[opencode.dispatch_steps]]
tool = "desktop_window_activate"
with = { hwnd = "{hwnd}" }
[[opencode.dispatch_steps]]
tool = "__sleep"
with = { ms = 500 }
[[opencode.dispatch_steps]]
tool = "desktop_input"
with = { hwnd = "{hwnd}", mode = "type", text = "{message}", send = "none" }
[[opencode.dispatch_steps]]
tool = "desktop_input"
with = { hwnd = "{hwnd}", mode = "hotkey", keys = ["enter"] }
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 350 lines · 0 tokens per session scan A d7f66500cdf3
agent-orchestration is a skill published in the GitHub repository mrpulor-gh/nuphus (61 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 7,007 tokens. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
tcapi
A helper for managing Tencent Cloud resources through Tencent Cloud's command-line API tool. Tencent Cloud is a cloud-services provider offering products such as virtual servers, storage, networks, and containers.
macos-harness
Control a whole Mac from one persistent Python session with screenshots, PID-targeted input, an animated virtual pointer, targeted Apple Accessibility, Apple Events, Browser Harness CDP, and filesystem access. Use for native, Electron, browser, dialog, file, or cross-app tasks without moving the physical cursor or…
chinese-video-transcribe-pdf
A workflow that turns Chinese-language videos, with or without subtitles, into structured text and a Chinese PDF report. It can use videos from sources such as YouTube or local MP4 files.
screen-use
A Windows desktop automation skill for clicking controls, entering text, using keyboard shortcuts, scrolling, and taking screenshots. It can also carry out multi-step tasks through visual guidance.
computer-use-macos
Top-level macOS computer-use skill with a bundled standalone runtime that bootstraps itself without any local Claude installation, private native modules, or extracted app assets.
control-window
Use when the user wants Claude to control, drive, or automate another app's window on macOS — clicking, typing, pressing keys, scrolling, or reading what's on screen in a window that isn't the terminal. Triggers on requests like "control my Telegram window", "drive this app for me", "click the button in that window"…